Entirely AI-generated and automatically published after automated validation. Not independently human fact-checked. Automated checks do not establish scientific accuracy.
This week’s clearest AI-for-materials update is a new MACE release adding magnetic models and other capabilities for atomistic modeling. The supplied records do not establish new experimental results or materials-discovery findings.
Featured: MACE v0.3.17 adds magnetic models and PolarMACE capabilities
MACE v0.3.17 adds MagneticMACE models for collinear and non-collinear magnetic moments, along with an ASE calculator. The release also adds opt-in polarizability and Fukui functions for PolarMACE, a MACE-MDP calculator, faster stress calculations, and lower memory use on large systems.
Why it matters: Magnetic models broaden MACE’s potential use in atomistic simulations of magnetic materials. The PolarMACE and performance changes may also support additional modeling workflows and larger calculations.
Limitations: These are release-note claims; the supplied evidence does not provide benchmarks or validate accuracy on specific materials. This is a software release, not evidence of experimental results. It requires Python 3.11+ and PyTorch 2.5+, and includes breaking changes.
v0.3.17 — Software release; Abstract only.